Vulnerability Discovery in Network Systems Based on Human-Machine Collective Intelligence

Abstract

Network vulnerability mining is an important topic in cyberspace security. Network vulnerabilities enable attackers to obtain sensitive information from computer systems, control computer systems illegally and cause severe damage. More effective vulnerability mining requires wider participation of cybersecurity engineers and intelligent computing devices as cooperation among the mining participants could take advantages of the complementarity capabilities of human and machine. The human and resource cost of network vulnerability mining can be remarkably reduced and the mining efficiency is improved accordingly. The principles and engineering mechanisms of introducing collective intelligence to network vulnerability mining is discussed in the paper and a vulnerability mining platform based on crowd intelligence is established following a four-layer system structure. Experimental tests showed that cooperation enables mining participants to work better and learn from empirical information, while better mining results could be obtained through procedure optimization.